体外膜肺氧合治疗急性呼吸窘迫综合征生存预测模型的构建  

Establishment of survival prediction model of extracorporeal membrane oxygenation in treatment of patient with acute respiratory distress syndrome

作　　者：王钊 李根 冯维静 吴澳 WANG Zhao;LI Gen;FENG Wei-jing;WU Ao(Department of Critical Care Medicine,Tianjin Third Central Hospital,Tianjin Key Laboratory of Extracorporeal Life Support for Critical Disease,Artificial Cell Engineering Technology Research Center,Tianjin Institute of Hepatobiliary Disease,Tianjin 300170,China)

机构地区：[1]天津市第三中心医院重症医学科,天津市重症疾病体外生命支持重点实验室,天津市人工细胞工程技术研究中心,天津市肝胆研究所,天津300170

出　　处：《生物医学工程与临床》2025年第1期78-86,共9页Biomedical Engineering and Clinical Medicine

基　　金：天津市医学重点学科(专科)建设项目资助(TJYXZDXK-025A)。

摘　　要：目的建立基于最小绝对收缩和选择算子(LASSO)惩罚的Logistic回归模型预测体外膜肺氧合(ECMO)治疗的成年人急性呼吸窘迫综合征(ARDS)患者生存结局。方法单中心、回顾性选择2009年1月至2022年12月于天津市第三中心医院接受ECMO治疗的重症ARDS成年人患者。预后随访时间截至2023年3月31日。共选择68例患者,其中男性54例,女性14例;年龄21~80岁,平均年龄46岁;身体质量指数(BMI)22.0~32.6 kg/m2,平均BMI 25.1 kg/m2;其中21例(30.8%)在ECMO辅助下完成跨院转运;急性生理与慢性健康评分Ⅱ(APACHEⅡ)得分8~44分,平均APACHEⅡ得分31分;Murray评分3~4分,平均Murray评分3.6分。根据患者血流动力学状况灵活选择两种ECMO模式,即静脉-静脉转流模式(v-v ECMO)和静脉-动脉转流模式(v-a ECMO)。ECMO辅助后逐渐降低呼吸机辅助条件实施“肺休息”策略。ECMO辅助期间常规给予“全覆盖”抗感染模式,对合并急性肾功能衰竭患者使用连续肾脏替代治疗,酌情补充血制品,视血流动力学指标变化微调正性肌力药物剂量,根据胃肠功能状况使用肠内、肠外营养治疗,对肺不张患者采取俯卧位通气,常规应用纤维支气管镜及支气管肺泡灌洗进行诊断性治疗。建立基于LASSO惩罚的Logistic回归模型预测患者生存结局。使用Kaplan-Meier生存曲线和受试者操作特性曲线评估模型的生存区分能力。结果ECMO建立前有创机械通气时长4.0~56.8 h,平均有创机械通气24.0 h。ECMO平均辅助(114.4±65.4)h,其中v-v模式55例(80.9%),v-a模式10例(14.7%),v-v转换为v-a模式3例(4.4%)。ECMO建立后机械通气时长105.6~259.3 h,平均机械通气时长164.0 h。重症监护病房和总住院时间分别为1~78 d[(15.3±19.0)d]、1~78 d[(20.6±19.3)d]。存活出院率为52.9%。单因素分析发现年龄、APACHEⅡ得分、ECMO前机械通气时长、重症监护病房停留时长、ECMO治疗后72 h心率、血乳酸、血小板计数、尿素氮、�Objective To establish Logistic regression model based on least absolute shrinkage and selection operator(LASSO)for predicting survival outcome of adult patients with acute respiratory distress syndrome(ARDS)performed extracorporeal membrane oxygenation(ECMO).Methods From January 2009 to December 2022(follow-up time was March 31 in 2023),a total of 68 adult patients with severe ARDS who received ECMO treatment were enrolled retrospectively in single center,which included 54 males and 14 females,aged 21-80 years old with mean age of 46 years old;body mass index(BMI)was 22.0-32.6 kg/m2 with mean BMI of 25.1 kg/m2.Among them,21 cases(30.8%)completed cross-hospital transport with ECMO-assisted;the acute physiology and chronic health evaluationⅡ(APACHEⅡ)score was 8-44 with mean APACHEⅡscore of 31.Murray score was 3-4 with mean score of 3.6.According to hemodynamic status of patient,2 ECMO modes were flexibly selected,namely venous-venous ECMO mode(v-v ECMO)and venous-arterial ECMO mode(v-a ECMO).After ECMO-assisted,the ventilator-assisted conditions were gradually reduced and“lung rest”strategy was implemented.During ECMO assisted period,the“full coverage”anti-infection mode was routinely given,the patients with acute renal failure were treated with continuous renal replacement therapy,blood products were supplemented as appropriate,the dose of positive inotropic drugs was fine-tuned according to changes of hemodynamic indexes,enteral and parenteral nutrition was used according to status of gastrointestinal function,prone position ventilation was adopted for patients with atelectasis.In addition,fiberoptic bronchoscopy and bronchoalveolar lavage were routinely used for diagnostic treatment.Logistic regression model based on LASSO penalty was established to predict survival outcome of patients.Kaplan-Meier survival curve and receiver operating characteristic curve were used to evaluate survival discrimination ability of model.Results The duration of invasive mechanical ventilation before establishment of

关 键 词：急性呼吸窘迫综合征 体外膜肺氧合 呼吸功能衰竭 预后 生存模型 

分 类 号：R563.8[医药卫生—呼吸系统]